Materials Synthesis and Processing

WILLIAM F.BRINKMAN

As the research discipline of materials science has matured, its theoretical base has broadened and become more rigorous. This progression has led to the development of new materials whose performance is vastly superior to that of only a few years ago. New techniques for materials synthesis draw on knowledge from physics, chemistry, and other disciplines to create entirely new materials. In addition, better theoretical understanding of established processes allows materials scientists to control materials properties to a degree previously unachievable. The following examples illustrate how better understanding of the fundamentals of materials synthesis and processing has led to improved, and in some cases new, materials.

These examples are (1) strained-layer superlattices—a new concept in semiconductor structures; (2) glass-ceramic materials—an older materials technology that has not been understood but holds promise for structural ceramics; and (3) gel-derived, controlled-porosity materials—solution-to-gelation chemical processing applied to antireflective coatings with tremendous potential for high-technology applications.

STRAINED-LAYER SUPERLATTICES

The strained-layer superlattice (SLS) is an important new field of semiconductor materials research. The SLS consists of many thin (a few tens of angstroms) layers of alternating, strained single-crystal semiconductor types (for example, GaAsP and GaAs). In structures of this type the lattice mismatch is accommodated totally by uniform lattice strain, and no misfit dislocations are generated at the interfaces (see Figure 1). This remains true as long as the layers are sufficiently thin (up to 250 angstroms for a 1.8 percent



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 307
Advancing Materials Research Materials Synthesis and Processing WILLIAM F.BRINKMAN As the research discipline of materials science has matured, its theoretical base has broadened and become more rigorous. This progression has led to the development of new materials whose performance is vastly superior to that of only a few years ago. New techniques for materials synthesis draw on knowledge from physics, chemistry, and other disciplines to create entirely new materials. In addition, better theoretical understanding of established processes allows materials scientists to control materials properties to a degree previously unachievable. The following examples illustrate how better understanding of the fundamentals of materials synthesis and processing has led to improved, and in some cases new, materials. These examples are (1) strained-layer superlattices—a new concept in semiconductor structures; (2) glass-ceramic materials—an older materials technology that has not been understood but holds promise for structural ceramics; and (3) gel-derived, controlled-porosity materials—solution-to-gelation chemical processing applied to antireflective coatings with tremendous potential for high-technology applications. STRAINED-LAYER SUPERLATTICES The strained-layer superlattice (SLS) is an important new field of semiconductor materials research. The SLS consists of many thin (a few tens of angstroms) layers of alternating, strained single-crystal semiconductor types (for example, GaAsP and GaAs). In structures of this type the lattice mismatch is accommodated totally by uniform lattice strain, and no misfit dislocations are generated at the interfaces (see Figure 1). This remains true as long as the layers are sufficiently thin (up to 250 angstroms for a 1.8 percent

OCR for page 307
Advancing Materials Research FIGURE 1 Strained-layer superlattice (SLS): schematic cross section showing substrate, buffer layer, and superlattice. mismatch). SLSs are typically made from the more common Group III–V compound semiconductors such as InAsSb, GaAsP, or InGaAs. The multiple thin layers may be viewed macroscopically as a new semiconductor material (material X). The combination of the thin layers and lattice strain allow flexibility in tailoring the properties of material X. The properties that may be controlled include the energy bandgap, direct-indirect bandgap, effective mass-velocity field characteristics, and average lattice constant. The energy bandgap of an SLS can be controlled by changing either the type of material or the thickness of the layers or both. The effect is akin to the behavior of the particle-in-box solution to the Schrödinger wave equation, that is, thin layers increase the energy ground state.

OCR for page 307
Advancing Materials Research The multiple thin layers also change the periodicity of the Brillouin zones. In some cases this can cause the valence band maximum and conduction band minimum to have the same crystal momentum in the SLS when they do not in the bulk crystal. The lattice constant of the SLS depends on the stiffness and thickness of the individual layers of the material. Hence, it is also possible to adjust the average lattice constant of this material. This greatly increases the number of choices for host substrates on which to grow these materials. The number of new semiconducting materials that can be made is almost infinite. Variations can be made in material type, layer thickness, and doping over a wide range. This tremendous variety dictates that researchers be able to synthesize their own SLSs to achieve significant progress in this field. Crystal growth by molecular beam epitaxy (MBE) is one technique that allows control of the growth of these materials to essentially monolayer levels. These new SLS materials can be synthesized to optimize material properties for a given device type, such as microwave, optoelectronic, and infrared (IR). An example of this is InAsSb SLS 8–12-micron IR detector material. None of the common Group III–V compound semiconductors has a bandgap small enough to reach the important 8–12-micron IR band. This has forced the development of HgCdTe detectors for this wavelength range. This material system has several undesirable material properties. Theoretical calculations have shown that the strains in an SLS can produce sufficient lattice dilatation in InAsSb to shift the absorption edge from 9 microns in the bulk alloy to greater than 12 microns in the SLS.1 This SLS is therefore an alternative to HgCdTe. Figure 2 shows a transmission electron micrograph of a typical InAsSb SLS. Another example of special semiconductors that can be synthesized only from SLSs is light-hole material. Many Group III–V compound semiconductors have electron effective masses much smaller than silicon. This results in n-type devices that are faster than similar silicon devices. However, the holes in most Group III–V semiconductors are degenerate, that is, the valence band contains both heavy and light holes. Hence, p-type devices made from Group III–V semiconductors are no faster than silicon devices. It has been shown both theoretically and experimentally that the strain and quantum well effect in specially designed SLSs in the InGaAs/GaAs system can cause a favorable splitting of this degeneracy and produce light-hole material for p-type devices.2 This development is important to high-speed, low-power logic circuits. Finally, it has been shown that the SLS can produce high-quality materials on lattice-mismatched substrates (defects are blocked by the SLS). This feature, coupled with the tunability of lattice constant and bandgap in SLSs, makes it possible to synthesize SLSs from terniaries (such as InGaAs) that can perform the same function as bulk quaternaries (such as InGaAsP). This

OCR for page 307
Advancing Materials Research FIGURE 2 Transmission electron micrograph of InAs0.2Sb0.8/InSb strained-layer superlattice for infrared detector applications. is important because it reduces the difficulty in fabricating devices such as semiconductor lasers at 1.5 microns. GLASS-CERAMIC MATERIALS A glass-ceramic is an inorganic material that is formed from molten glass and is subsequently transformed to a polycrystalline ceramic object by con-

OCR for page 307
Advancing Materials Research trolled crystallization. Glass-ceramics are used in applications as diverse as dinnerware, telescope mirrors, radomes, and semiconductor substrates because frequently they are more economical to produce and because they possess special properties that cannot be achieved by other means. Control of the crystallization process is the key to obtaining desirable properties in those materials. Most glass-ceramic formulations contain small amounts of special additives, called nucleating agents, that initiate the crystallization process and influence the particular mix of phases that develops. Platinum, TiO2, ZrO2, and P2O5 in concentrations from 0.01 to 10 percent are commonly used nucleating agents in silicate-based glass-ceramics. Although the nucleation and growth process in glass-ceramics has been extensively studied in the 25 years since it was discovered, at present there is no general theory that explains how the nucleating agents operate. Most available models are specific to a given system, and they commonly postulate the formation of some sort of heterogeneity by the nucleating agent that catalyzes the subsequent crystallization.3 Recently, the nucleation and crystal growth mechanism in a P2O5-doped lithium silicate glass-ceramic has been explained in detail.4 Transmission electron microscopy, x-ray diffraction, and differential thermal analysis showed that at the nucleation temperature, the P2O5 reacts with the glass constituents to precipitate small Li3PO4 crystallites throughout the molten glass. The size and degree of faceting of the Li3PO4 crystallites increases with time at the nucleation temperature (Figure 3). During crystallization, desired phases of lithium metasilicate, lithium disilicate, and cristobalite grow epitaxially on preferred {120} and {010} faces of the Li3PO4. There is a definite crystallographic orientation relationship between the Li3PO4 substrate and the different crystal phases, with a d-space mismatch of less than 5 percent (Figure 4). Further research revealed that the Li3PO4 crystallites must grow to some critical size before their {120} faces are sufficiently faceted to serve as the substrates for the epitaxial growth (Figure 3). The concentrations of the different crystalline phases can be varied reproducibly using different nucleation times to change the size of {120} faces on the Li3PO4 crystallites. This work is important because it is the first direct observation of a nucleation mechanism that was hypothesized more than 20 years ago. Further, the specific orientation relationships that are observed may allow the existing qualitative theory to be made quantitative. The discovery that the phase concentration is controlled by the size of preferred faces in the nucleant is unprecedented. That fact, and our understanding of the nucleation mechanism, allows us to vary the expansion coefficient of the resulting glass-ceramic reproducibly from 100 to 160×10–7/°C. One glass-ceramic in that family that has an expansion coefficient of 145×10–7/°C is now used to make high-strength glass-ceramic-to-metal seals with Inconel 718 for pyrotechnic actuators, connectors, and other electrical components. The under-

OCR for page 307
Advancing Materials Research FIGURE 3 Growth of lithium phosphate crystallites at 1,000°C, in which {120} faces develop. standing of the crystallization mechanism permits close control of the glass-ceramic expansion coefficient (Figure 5), which in turn minimizes materials stresses so that the resulting components are more than three times stronger than those they replaced. GEL-DERIVED, CONTROLLED-POROSITY MATERIALS: ANTIREFLECTIVE COATINGS Chemical syntheses and powder processing in solution are used at Sandia National Laboratories to produce a variety of technologically important in-

OCR for page 307
Advancing Materials Research FIGURE 4 Crystallization of a glass-ceramic by epitaxial growth. From: T.J.Headley and R.E. Loehman, Sandia National Laboratories.

OCR for page 307
Advancing Materials Research FIGURE 5 (a) Thermal expansion coefficient as a function of time at 1,000°C; (b) cristobalite concentration as a function of time at 1,000°C. organic materials—for example, electronic and photoelectronic ceramics, catalyst substrates, and refractory glasses. One area of intense interest that has developed in the last five years is the solution-to-gelation route to inorganic glasses—the sol-gel process. Sandia has an extensive R&D program to understand and exploit this important technology. The sol-gel process uses monomeric compounds (normally alkoxides) of network-forming elements (for example, silicon, aluminum, titanium, and

OCR for page 307
Advancing Materials Research boron) as glass precursors. In alcoholic solutions catalyzed by the addition of acids or bases, the alkoxides are hydrolyzed and condensed to form inorganic oxide networks composed of—M—O—M—linkages as follows: Si(OR)4+xcH2O→Si(OH)x(OR)4–x+xROH (1) n[Si(OH)x(OR)4–x]→nSiO2+n(x–2)H2O+n(4–x)ROH. (2) By control of growth conditions or precursor chemistry, the topology of the condensed species can be varied from dense balls of anhydrous oxide to wispy, ramified structures (random fractals). This control makes it possible to tailor the microstructure to a particular application,5 as seen in Figure 6. FIGURE 6 Various ways of controlling structural properties of materials during growth in solution.

OCR for page 307
Advancing Materials Research Inorganic oxides exhibiting controlled pore size, pore volume, refractive index, and chemistry (for example, acidity/basicity) are of interest in high-technology applications such as catalysts, sensors, and photonics. We describe here one application of sol-gel-derived controlled-porosity materials: quarter-wave interference films used to make glasses and plastics antireflective for solar and inertial confinement fusion applications.6 In work at Sandia, controlled growth in solution has been used to tailor the refractive index of gel-derived SiO2 thin films. Working in a pH and concentration regime where growth occurs by a process analogous to random agglomeration, we observe that the hydrodynamic radius of the scattering species increases with t1/2 and the refractive index of the corresponding deposited films decreases with t1/2 (Figure 7). For a single-layer antireflective film, the reflectance at a specific wavelength will be zero when the refractive index of that film, nf, is given by nf=(ngna)1/2, (3) where ng is the refractive index of the substrate and na is the refractive index of the ambient atmosphere. The very low refractive indices required in antireflective films for vitreous silica, lucite, and polycarbonates (nf≅1.22–1.25) are easily achieved by this controlled-growth method. To understand this process, it is necessary to consider the structures of solution-grown polymers and aggregates. In dilute solution near the isoelectric point of SiO2 (neutral charge), growth occurs as primary species randomly diffuse in a growing embryo.7 As shown in Figure 6 this results in fractal structures, that is, structures in which the density decreases radially from the center of mass (incremental pore volume increases radially from the center of mass). The refractive index of thin films is an inverse function of pore volume. The fact that the refractive index of deposited films decreases (pore volume increases) with the hydrodynamic radius is due to the fractal nature of the solution-grown polymers. Whereas dense or uniformly porous species should randomly close-pack to the same pore volume regardless of size, packing of fractal objects that do not interpenetrate causes the pore volume to increase (refractive index decreases) with the size of the primary aggregate (Figure 6). This concept allows us to deposit a quarter-wave interference film on lucite using a one-step process requiring no heat treatment or acid etching. Single-layer interference films exhibit discrete minima in reflectance as a function of wavelength. This minimum occurs when the optical thickness, nfd, is an odd multiple of one-quarter of the wavelength of the incident light. Thus, by adjustment of the film thickness, d, we are able to minimize reflectance at any desired wavelength, for example, at the solar spectrum maximum, 580 nm.

OCR for page 307
Advancing Materials Research FIGURE 7 Interrelationship between polymer growth and refractive index. The process of controlled growth in solution to obtain controlled-porosity materials is a general technique adaptable to numerous applications. Sandia has applied for a patent covering these concepts. ACKNOWLEDGMENT The author would like to acknowledge the help of R.Chafin, R.Quinn, and T.J.Headley in preparing this chapter.

OCR for page 307
Advancing Materials Research NOTES 1.   G.C.Osbourn, J. Vac. Sci. Technol. B2 (3), 176 (1984). 2.   J.E.Schirber, I.J.Fritz, and L.R.Dawson, Appl. Phys. Lett. 46 (2), 187 (1985). 3.   P.W.McMillan, Glass Ceramics, 2nd ed. (Academic Press, New York, 1979). 4.   T.J.Headley and R.E.Loehman, J. Am. Ceram. Soc. 67 (9), 620–625 (1984). 5.   C.J.Brinker, G.W.Scherer, and E.P.Roth, J. Non-Cryst. Solids 70, 301 (1985); 72, 345, 369 (1985). 6.   R.B.Pettit and C.J.Brinker, Soc. Photo-opt. Instrum. Eng., 562, paper no. 33 (1985). 7.   A.J.Hurd, in Proceedings of the International Symposium on Physics of Complex and Super-Molecular Fluids: Colloids, Micelles and Micro-Emulsions, Exxon Monograph Series, edited by S.Safron and N.Clark (Wiley, New York, in press).